mirror of
https://github.com/pre-commit/action.git
synced 2025-11-10 03:56:55 +00:00
808ec6adfe
These steps don't make use of any extra third-party actions so should hopefully be a safe suggestion for other people to use, and more efficient than getting `actions/checkout` to clone the entire Git history, which can take a long time on large repositories.
109 lines
3.2 KiB
Markdown
109 lines
3.2 KiB
Markdown
[](https://results.pre-commit.ci/latest/github/pre-commit/action/master)
|
|
[](https://github.com/pre-commit/action/actions)
|
|
|
|
pre-commit/action
|
|
=================
|
|
|
|
a GitHub action to run [pre-commit](https://pre-commit.com)
|
|
|
|
### using this action
|
|
|
|
To use this action, make a file `.github/workflows/pre-commit.yml`. Here's a
|
|
template to get started:
|
|
|
|
```yaml
|
|
name: pre-commit
|
|
|
|
on:
|
|
pull_request:
|
|
push:
|
|
branches: [master]
|
|
|
|
jobs:
|
|
pre-commit:
|
|
runs-on: ubuntu-latest
|
|
steps:
|
|
- uses: actions/checkout@v2
|
|
- uses: actions/setup-python@v2
|
|
- uses: pre-commit/action@v2.0.0
|
|
```
|
|
|
|
This does a few things:
|
|
|
|
- clones the code
|
|
- installs python
|
|
- sets up the `pre-commit` cache
|
|
|
|
### using this action with custom invocations
|
|
|
|
By default, this action runs all the hooks against all the files. `extra_args`
|
|
lets users specify a single hook id and/or options to pass to `pre-commit run`.
|
|
|
|
Here's a sample step configuration that only runs the `flake8` hook against all
|
|
the files (use the template above except for the `pre-commit` action):
|
|
|
|
```yaml
|
|
- uses: pre-commit/action@v2.0.0
|
|
with:
|
|
extra_args: flake8 --all-files
|
|
```
|
|
|
|
### only check files that have changed
|
|
|
|
`actions/checkout@v2` does a shallow clone by default which means pre-commit
|
|
doesn't have enough Git history to determine which files have changed in order
|
|
to only check them. So you need to expand the history yourself after the
|
|
checkout action has run.
|
|
|
|
This can be done effectively by getting a blobless clone with:
|
|
```yaml
|
|
- uses: actions/checkout@v2
|
|
- name: Expand Git history
|
|
run: |
|
|
# Tell Git we want a blobless clone,
|
|
# which excludes file contents of historical commits making the cloning much faster.
|
|
git config remote.origin.partialclonefilter blob:none
|
|
|
|
# Then fetch the history of the repository, skip tags since we don't need them.
|
|
git fetch --unshallow --no-tags
|
|
```
|
|
|
|
and you can then configure pre-commit to run with:
|
|
```yaml
|
|
- uses: pre-commit/action@v2.0.0
|
|
with:
|
|
extra_args: --from-ref ${{ github.event.pull_request.base.sha }} --to-ref HEAD
|
|
```
|
|
|
|
### using this action in private repositories
|
|
|
|
this action also provides an additional behaviour when used in private
|
|
repositories. when configured with a github token, the action will push back
|
|
fixes to the pull request branch.
|
|
|
|
using the template above, you'll make two replacements for individual actions:
|
|
|
|
first is the checkout step, which needs to use unlimited fetch depth for
|
|
pushing
|
|
|
|
```yaml
|
|
- uses: actions/checkout@v2
|
|
with:
|
|
fetch-depth: 0
|
|
```
|
|
|
|
next is passing the token to the pre-commit action
|
|
|
|
```yaml
|
|
- uses: pre-commit/action@v2.0.0
|
|
with:
|
|
token: ${{ secrets.GITHUB_TOKEN }}
|
|
```
|
|
|
|
note that `secrets.GITHUB_TOKEN` is automatically provisioned and will not
|
|
require any special configuration.
|
|
|
|
while you could _technically_ configure this for a public repository (using a
|
|
personal access token), I can't think of a way to do this safely without
|
|
exposing a privileged token to pull requests -- if you have any ideas, please
|
|
leave an issue!
|